The current issue and full text archive of this journal is available on Emerald Insight at:

www.emeraldinsight.com/0143-7739.htm

LODJ
37,6
Leadership: is bad stronger
than good?
Maria Fors Brandebo and Sofia Nilsson
690 Department of Security, Strategy and Leadership,
Swedish Defence University, Karlstad, Sweden, and
Received 24 September 2014 Gerry Larsson
Revised 17 April 2015
25 August 2015
Department of Security, Strategy and Leadership,
27 August 2015 Swedish Defence University, Karlstad, Sweden and
Accepted 31 August 2015
Department of Public Health, Hedmark University College,
Elverum, Norway

Abstract
Purpose – The purpose of this paper is to investigate if the thesis “bad is stronger than good” also
holds true for a number of leadership issues, more specifically: trust in the immediate leader, emotional
exhaustion, work atmosphere and propensity to leave.
Design/methodology/approach – Questionnaire responses were obtained from military personnel
in Estonia, Sweden, Switzerland and the Netherlands (n ¼ 625).
Findings – Multiple regression analyses revealed a certain pattern. Constructive leadership
behaviours showed stronger positive associations with trust in the immediate supervisor and work
atmosphere, than destructive leadership behaviours showed negative associations. On the other
hand, destructive leadership behaviours showed stronger positive associations with emotional
exhaustion and propensity to leave, than constructive leadership behaviours showed negative
associations. This suggests that constructive leadership behaviours possibly have a greater
impact on positive phenomenon and/or phenomenon associated with work-related relationships.
On the other hand, destructive leadership behaviours appear to have a greater impact on negative
phenomena with a stronger personal meaning. The results also show that the passive forms of
destructive leadership are the behaviours that had the strongest impact on the investigated
dependent variables.
Research limitations/implications – Limitations related to item construction, common method
variance, response set tendencies, translation of the instruments, and lack of response rate are
discussed.
Practical implications – The results emphasize the importance of focusing on both constructive and
destructive leadership at the selection stage, as well as during training of military leaders. Focusing on
them separately obstructs optimal leader development and prevents leaders from gaining authentic
self-knowledge. The results also point at the importance of including both aspects of leadership in
leader evaluation processes.
Originality/value – The use of both constructive and destructive leadership behaviours with
respondents from multiple nations in the same analysis.
Keywords Military, Trust, Emotional exhaustion, Constructive leadership, Destructive leadership,
Propensity to leave, Work atmosphere
Paper type Research paper

Leadership & Organization

Development Journal The authors thank Leo Kant for his comments on an earlier draft of the paper. The authors also
Vol. 37 No. 6, 2016
pp. 690-710 thank Miriam C. de Graaff, MSc, Dutch Ministry of Defence, Merle Parmak, PhD, Estonian
© Emerald Group Publishing Limited
0143-7739
National Defence College and Andres Pfister, PhD, Military Academy at ETH Zurich for your
DOI 10.1108/LODJ-09-2014-0191 collaboration and for collecting data for the present study.
Introduction Leadership: is
The hypothesis that “bad is stronger than good” has been suggested to be general bad stronger
across a broad range of psychological and social phenomena (Baumeister et al., 2001),
including leadership (Einarsen et al., 2007). This means that negative events will tend to
than good?
have a greater impact on the individual than positive events of the same type, and good
can only prevail over bad by superior force in numbers (Baumeister et al., 2001). In line
with this, an increasing number of leadership researchers are recognizing the “dark 691
sides of leadership”. Hence, both the negative and positive aspects of the relationship
between leaders and followers are being considered (Clements and Washbush, 1999;
Einarsen et al., 2007). Studies across several work environments reveal that there is
often a strong prevalence of destructive leadership behaviours (Aasland et al., 2010;
Glasø et al., 2010). Focusing only upon constructive leadership behaviours might
therefore limit the understanding of actual influence processes between leaders and
subordinates, which are composed of both negative and positive aspects.
Einarsen et al. (2007) have defined constructive leadership as acting in “accordance
with the legitimate interests of the organisation, supporting and enhancing the goals,
tasks, and strategy of the organisation, as well as making optimal use of organisational
resources. Simultaneously, they enhance the motivation, well-being and job satisfaction
of their followers by engaging in behaviours such as inviting subordinates to
an extended engagement, and granting involvement and participation in decision
processes. These leaders are concerned with the welfare of their subordinates while
simultaneously being focused on goal attainment and the effective use of resources in
the service of the legitimate interests of the organisation” (p. 214). There are
several constructive leadership constructs (e.g. authentic leadership and transactional
leadership) of which the most used and well researched is transformational leadership.
Transformational leadership aims to build trust, admiration, loyalty and respect from
the subordinates towards the leader, resulting in the subordinates becoming motivated
to perform more than they originally set out to do (Yukl, 2006). In this paper, we use
developmental leadership to define constructive leadership. Developmental leadership
is a Scandinavian development of transformational leadership. The main differences
between transformational leadership and developmental leadership are that the latter is
grounded in an interactionistic person-by-situation base. Charismatic leadership has
also been toned down since charismatic leaders, in a Scandinavian context, are often
associated with less favourable leaders (such as Hitler) (Larsson et al., 2003).
Destructive leadership has been defined as “the systematic and repeated behaviour
by a leader, supervisor or manager that violates the legitimate interest of the
organisation by undermining and/or sabotaging the organisation’s goals, tasks,
resources and effectiveness and/or the motivation, well-being or job satisfaction of
subordinates” (Einarsen et al., 2007, p. 208). Krasikova et al. (2013) modified the
definition by Einarsen et al. (2007) by: arguing that destructive leadership should be
viewed as harmful behaviour imbedded in the process of leading (and by excluding
behaviours falling under counterproductive work behaviour); distinguishing between
encouraging subordinates to pursue destructive goals and using destructive methods
of influence with subordinates; and defining destructive leadership as volitional
behaviour. Schyns and Schilling (2013) propose another definition of destructive
leadership: “A process in which over a longer period of time the activities, experiences
and/or relationships of an individual or the members of a group are repeatedly
influenced by their supervisor in a way that is perceived as hostile and/or obstructive”
(p. 141). This definition does not include the anti-organization dimension but rather
LODJ focuses on the result of and the subordinates’ perception of the leader’s behaviour.
37,6 Similar to Einarsen et al. (2007), this definition also stresses that the behaviour is
repeated over time. Consequently, occasional use of destructive leadership behaviours
are not considered destructive leadership. This study takes its point of departure in the
definition proposed by Einarsen et al. (2007), viewing destructive leadership as
systematic and repeated behaviour that can be directed both towards the subordinates
692 and the organization.
Some scholars divide destructive leadership behaviours into active and passive
forms (Einarsen et al., 2007; Larsson et al., 2012). While active forms represent more
deliberate and volitional behaviours, passive forms are regarded as behaviours leaders
use when they have more or less abdicated from supervisor responsibilities and duties
(Einarsen et al., 2007). Researchers have discussed whether or not laissez-faire
leadership (non-leadership, inactive) should in fact be regarded as a form of destructive
leadership. There appear to be diverging opinions: some argue that a concept should
not be defined by its consequences and that laissez-faire leadership is more related to
ineffective behaviours than to destructive ones (Craig and Kaiser, 2012; Krasikova
et al., 2013; Schyns and Schilling, 2013). In contrast, others have called attention to the
negative consequences of passive, indirect leader behaviours on, for example,
job satisfaction, efficiency, workplace stressors, bullying at work and psychological
distress (Frischer and Larsson, 2000; Kelloway et al., 2005; Neuman and Baron, 2005;
Skogstad et al., 2007). This has led to the conclusion that laissez-faire leadership is a
form of destructive leadership. Skogstad et al. (2014) state that this kind of leadership
can be defined as a follower-centred form of avoidance-based leadership and is thus
perceived as a volitional and active avoidance of subordinates when they are in need of
leadership and support. Laissez-faire leadership has also been highlighted as the most
prevalent destructive leadership behaviour (Aasland et al., 2010). Furthermore,
two recently developed scales measuring destructive leadership include laissez-faire
leadership behaviours (Larsson et al., 2012; Thoroughgood et al., 2012). This paper
supports the idea that laissez-faire leadership is a form of destructive leadership.
We agree with the opinion of Skogstad et al. (2007, 2014) that laissez-faire leadership
appears to be more of a volitional, active and counterproductive leadership style than a
zero type/ineffective. We also argue that a leader can intentionally be absent and
uncommitted to his/her subordinates. A leader who chooses to prioritize the pursuit of
their own interests or merits above the task of staff management may do so
intentionally. Nevertheless, it may be difficult for subordinates to discover if the
leader’s absence is due to intentional priorities of self-interest or to ineffectiveness.
In this respect, it is difficult to say whether or not a passive form of leadership is in fact
destructive behaviour and it may instead be regarded as ineffective/non-leadership.
However, if the outcomes of a leadership are destructive for the organization and/or the
subordinates, the behaviour should be regarded as destructive regardless of the cause
(inefficiency, etc.) or intention, especially if the behaviour is repeated and systematic.
The growing interest in destructive leadership is suggested to be related to its costs,
as it is noted that destructive leadership leads to absenteeism, turnover and impaired
effectiveness (Schyns and Schilling, 2013). Although researchers argue that bad is
stronger than good (Baumeister et al., 2001; Einarsen et al., 2007), a recent meta-analysis
revealed contradictory results, suggesting that bad may not always be stronger
regarding leadership (Schyns and Schilling, 2013). Constructive leadership was found
to have higher positive correlations than destructive leadership’s negative correlations,
with different outcomes such as job satisfaction, attitude towards the leader, turnover
intention and individual performance. The only exceptions were commitment and Leadership: is
well-being, where destructive leadership showed higher correlations (Schyns and bad stronger
Schilling, 2013). Only a few studies have included both constructive and destructive
leadership behaviours in the same analysis. However, most are limited in the sense that
than good?
they, for example, only include passive forms of destructive leadership (Mullen et al., 2011)
or social support/undermining behaviours (Duffy et al., 2002). In the present study, we
therefore take a broader approach and examine if destructive or constructive leadership 693
behaviours are the best predictors of follower work outcomes.
Follower outcomes of destructive leadership are of special study interest. In a military
context, destructive leadership is brought to a head, as armed forces often perform tasks
in life or death situations, putting great strain on the health and well-being of military
personnel (Sweeney et al., 2011). Military teams frequently work closely together for long
periods of time and under demanding conditions. In order to solve the task, team
members need to get along and be able to collaborate. Individuals who are unable to
handle the stress and pressure are not only a risk to themselves, but to the whole group
(Campbell, 2012; Ness et al., 2011; Sweeney et al., 2011). Below follows a more thorough
description of those outcomes as observed in the available research.
The aim of this study was to investigate if the thesis “bad is stronger than good”
also holds true for a number of leadership outcomes, more specifically: trust in the
immediate leader, work atmosphere, emotional exhaustion and propensity to leave.

Trust in the leader

A number of scholars identify trust in leaders as a crucial factor in both civilian and
military contexts (Belenky et al., 1985; Kolditz, 2007; Sweeney et al., 2009; Sweeney, 2010).
Kolditz (2007) has claimed that leadership in environments such as the military are more
dependent on trust. Military subordinates are expected to give up their right to
self-determination and trust the leader (follow orders) (Collins and Jacobs, 2002).
Lack of trust in the leader may thus lead to negative consequences in the form of severe
injuries or death. The strongest effects of destructive leadership have been found to be
related to how followers feel about their leader (attitudes towards their leader)
(Schyns and Schilling, 2013). Especially Tepper’s (2000) abusive supervision scale has
been found to be more strongly related to attitudes and behaviours directed towards
the leader compared to outcomes such as turnover intention and performance.
However, Schyns and Schilling (2013) suggest that this may be due to item content since
Tepper’s (2000) concept of abusive supervision is claimed to have a stronger personal
meaning (for instance affectivity and well-being) than other concepts that are more work
related (for instance turnover intention and performance). While Kelloway et al. (2005)
found that constructive and destructive forms of leadership had approximately the same
effect on trust in the leader, the meta-analysis of Schyns and Schilling (2013) revealed
higher correlations for constructive leadership and attitudes towards the leader than for
destructive leadership. Contrary to “bad is stronger than good” we therefore suggest that:
H1. Constructive leader behaviours show stronger positive associations with
trust in the immediate supervisor than destructive leader behaviours show
negative associations.

Work atmosphere
Soldiers and officers often work closely together under stressful and demanding
conditions. In order to maintain one’s spirits and be capable of functioning, a good work
LODJ atmosphere is of utmost importance. Work atmosphere has not been particularly
37,6 highlighted as an outcome of destructive leadership. However, research shows that
passive leadership may create a work climate characterized by frustration and stress,
resulting in anti-social behaviour in work groups (Einarsen, 1999; Leymann, 1996).
Laissez-faire leadership has also shown positive correlations with experiences of
workplace stressors such as role conflicts and role ambiguity, leading to conflicting
694 expectations. Further, laissez-faire leadership has been found to have a high correlation
with high conflict levels in working groups, which can be assumed to affect the work
atmosphere in the organization (Hinkin and Schriesheim, 2008; Skogstad et al., 2007).
Previous research also suggests that the relationship between laissez-faire leadership
and workplace stressors are mainly explained by the presence of laissez-faire
leadership, not by the lack of constructive leadership (Skogstad et al., 2007). Given this,
we assume that:
H2. Destructive leadership behaviours show stronger negative associations
with workplace atmosphere than constructive leadership behaviours show
positive associations.
H3. Passive destructive leadership behaviours show stronger negative associations
with workplace atmosphere than active destructive leadership behaviours.

Emotional exhaustion
In military contexts, mental endurance and strength are paramount to preserving
battle fitness, making this an important issue to investigate. The relationship between
destructive leadership and individual-related consequences such as stress
and well-being is well documented (Burris et al., 2008; Chen and Kao, 2009; Schyns
and Schilling, 2013; Tepper, 2000). To illustrate, destructive leadership has been
suggested to lead to subordinates expressing emotional exhaustion (Chi and Liang,
2013; Harvey et al., 2007; Tepper, 2000). This may have negative organization-related
consequences since individuals who feel emotionally exhausted tend to withdraw from
their job in order to prevent further resource depletion (Chi and Liang, 2013). Based on
the principle that bad is stronger than good, and the meta-analytical finding that
reveals higher correlations between destructive leadership and well-being compared to
constructive leadership and well-being (Schyns and Schilling, 2013), we therefore
suggest that:
H4. Destructive leadership behaviours show stronger positive associations with
subordinates’ emotional exhaustion than constructive leadership behaviours
show negative associations.

Propensity to leave the profession

Several nations have abandoned their conscript systems in favour of an all-volunteer
force. When all personnel apply voluntarily to the armed forces, greater demands are
put on leaders to motivate their subordinates to continue their employment (Nilsson
et al., 2012). Recruiting and educating new individuals can be costly and have negative
consequences on the effectiveness of the organization. Several studies have emphasized
the influence of the immediate leaders on the propensity of subordinates to leave the
organization ( Jones et al., 1996; Larsson et al., 2012). A lack of transformational
leadership has been shown to have an influence (although modest) on nurses’
propensity to leave their profession (Bycio et al., 1995). Research on destructive Leadership: is
leadership has also revealed that negative leader behaviours are related to higher bad stronger
degrees of turnover (Ashforth, 1994; Schyns and Schilling, 2013). For example,
Larsson et al.’s (2012) findings showed a positive correlation between destructive
than good?
leadership and propensity to leave the armed forces regarding conscripts and cadets.
However, the study did not reveal a corresponding relationship for the higher-level
officers (majors) in the study. Other studies have also shown that destructive 695
leadership does not have a negative impact on military officers’ inclination to remain in
service (Reed and Bullis, 2009). Reed and Bullis’ (2009) study only included higher-rank
officers. We therefore suggest that:
H5. Constructive leadership behaviours show stronger negative associations with
subordinates’ propensity to leave the organization than destructive leader
behaviours show positive associations.
A summary of the five hypotheses is presented in Table I.

Method
Participants and procedure
Questionnaire responses were obtained from military personnel in Estonia, Sweden,
Switzerland and the Netherlands attending courses at their respective national defence
academies. Before giving informed consent, all participants received oral and/or written
information about the study. The Estonian and Dutch data were collected in two ways:
during classroom settings; and online (by e-mail). The Swedish and Swiss data were
collected during ordinary classroom settings. Participants responded anonymously.
The response rate was 100 per cent for the Estonian sample, 90 per cent for the Dutch
sample and 62 per cent for the Swedish sample. The response rate for the Swiss sample
is unknown. The number of participants was originally 625. Due to handling of missing
values, the final number of participants was 533. The handling of missing values was
conducted as follows. On the main instrument, the Destrudo-L, participants with two or
more missing values on each of the four-item factor scales were dropped. For those with

Outcome measures
Leadership Emotional
behaviour Trust Work atmosphere exhaustion Propensity to leave

Constructive Stronger positive Stronger negative

(C) association for C association for C
than negative than positive
association for D association for D
(H1) (H5)
Destructive (D) Stronger negative Stronger positive
association for D association for D than
than positive negative association
association for C for C (H2)
(H4)
Addendum, stronger
negative association Table I.
for passive D than for Summary of study
active D (H3) hypotheses
LODJ one out of four missing, the individual mean of the three obtained scores was entered as
37,6 the score on the item with a missing value. As this procedure was quite labour intense,
on all other scales, missing values were replaced with the overall mean score on the
item in question. A few additional participants were omitted from the regression
analyses as they were deemed as outliers (W 3.0 SD). A description of the national
samples is shown in Table II.
696 Table II shows that the proportion of men and women differs significantly across
the four countries. The Estonian and the Swedish samples have a larger proportion of
women, while the proportion of men is higher in the Netherlands and Switzerland.
The Swiss and the Swedish samples have a larger proportion of individuals aged 29 or
younger, while the proportion of individuals aged 30 or older is higher among the
participants from Estonia and the Netherlands. The Dutch and Swedish samples have a
larger proportion of individuals with higher education levels. The Swedish sample also
has a larger proportion of individuals working in the army compared to Estonia and
Switzerland. In the Swiss sample, none of the rated leaders were female.
The participants from the Netherlands and Sweden also report the most favourable
mean scores on the personality scale emotional stability.

Measures
The questionnaire was originally developed in Swedish and translated into English.
Contacts in each participating country translated the questionnaire from English into
their respective native language.
Emotional stability. Data were collected using the one item from the single-item
measure of personality (Woods and Hampson, 2005) designed to measure emotional
stability (neuroticism reversed) dimension/factor in the Big Five model of personality
(Costa and McCrae, 1992): how much does each description sound like you? Generally,
I come across as: someone who is sensitive and excitable, and can be tense. Someone
who is relaxed, unemotional, rarely gets irritated and seldom feels blue. The factor/item
is measured on a nine-point, bipolar graded line.
Knowledge of the leader. Knowledge of the leader was measured using the item
“I know my immediate supervisor/commander this well”. The scale ranged from 1,
“Not at all” to 6, “Very well”.
Destructive leadership behaviours. The Destrudo-L, a 20-item questionnaire developed
by Larsson et al. (2012) was used to measure destructive leadership behaviours.
The questionnaire consists of five factors with four items in each: arrogant, unfair (sample
item “Treats people differently”), Cronbach α: 0.75; threats, punishments, overdemands
(sample item “Uses threats to get his/her way”), Cronbach α: 0.69; ego oriented, false
(sample item “Does not keep promises”), Cronbach α: 0.78; passive, cowardly (sample item
“Does not show an active interest”), Cronbach α: 0.75; and uncertain, unclear, messy
(sample item “Is bad at structuring and planning”) Cronbach α: 0.78.
Constructive leadership behaviours. To measure constructive leadership behaviours,
21 items forming three factors (exemplary model, individualized consideration, and
inspiration and motivation) from the Developmental Leadership Questionnaire (DLQ)
(Larsson, 2006) were used. As no specific hypothesis was developed regarding the three
factors, a mean score based on all 21 items was used. The Cronbach α coefficient for
the scale was 0.96. Sample items: “Acts in accordance with the opinions he or she
expresses”, “Takes time to listen” and “Creates enthusiasm for a task”. The response
scale on the original questionnaire ranges from 1 to 9. However, in order to be able to
 Estonia The Netherlandsb Switzerland Sweden
n ¼ 41 n ¼ 319 n ¼ 68 n ¼ 105
Background characteristics n % M SD n % M SD n % M SD n % M SD χ2/F pa

Gender
Women 5 12 14 4 3 4 15 14
Men 36 88 305 96 65 96 89 86 14.63 0.002
Age
p29 years 4 10 53 17 28 41 36 35
X30 years 37 90 266 83 40 59 68 65 32.25 0.000
Civilian education
Secondary school/high school 8 20 180 56 18 27 45 43
College/university 33 80 139 44 50 73 59 57 35.88 0.000
Branch of service
Army 30 73 52 77 43 42
Navy 1 2 1 2 30 29
Air force 9 22 11 16 29 28 47.22 0.000
Sex of rated leader
Male 47 90 314 98 68 100 97 92
Female 4 10 5 2 0 0 8 8 17.36 0.001
Emotional stability 5.37 1.51 6.45 1.71 5.55 1.69 6.39 1.89 0.000
Knowledge of rated leader 3.61 1.12 3.68 1.19 4.03 1.18 3.79 1.24 ns
Notes: aA Bonferroni correction was applied. All differences are at a 0.008 level of significance; bInformation on branch of service is lacking for the Dutch
sample

the participants
bad stronger
than good?

Basic description of
Table II.
697
Leadership: is
LODJ mix items from Destrudo-L and DLQ on the same scoring sheet, the DLQ response scale
37,6 was modified. Thus, the response scale on all items ranged from 1, “Never or almost
never” to 6, “Very often or always”.
Trust. Trust was measured with two items: “How high is your trust in your
superior/commander’s individual characteristics?” and “How high is your trust in
your superior/commander as a leader?” The items were formulated based on a study
698 on trust in military leaders (Fors Brandebo and Larsson, 2012). The scale ranged
from 1, “low” to 6, “high”. The Cronbach α coefficient for this scale was 0.91.
Work atmosphere. Work atmosphere was measured with four items based on a
previous study on military personnel’s motivation to continue in the armed forces
(Larsson et al., 2007). Sample items included: “I’m getting on well with my colleagues”
and “I feel that we have a nice atmosphere at work”. The scale ranged from 1, “Do not
agree” to 6, “Agree totally”. The Cronbach α coefficient for the scale was 0.78.
Emotional exhaustion. In order to measure emotional exhaustion, nine items from
the Maslach Burnout Inventory (Maslach et al., 1996) was used. Sample items included:
“I feel emotionally drained from my work” and “Working with people all day is really a
strain for me”. The respondents estimated the frequency on a scale from 1, “a few times
a year” to 6, “every day”. The Cronbach α coefficient for the scale was 0.88.
Propensity to leave the profession. Propensity to leave the armed forces was
measured with four items based on a previous study on military personnel’s motivation
to continue in the armed forces (Larsson et al., 2007). Sample items included: “I have
plans to change employer”, and “I will quit shortly”. The scale ranged from 1, “Do not
agree” to 6, “Agree totally”. The Cronbach α coefficient for the scale was 0.73.

Statistics
Dimensional analysis of the 20 destructive leadership behaviour items based on the
covariance matrix was performed using structural equation modelling (SEM) with
maximum likelihood estimates. The software Amos was used. The five-factor model
obtained in the original Destrudo-L (Larsson et al., 2012) was used as point of departure.
Acceptable model fit was determined at a root mean square error of approximation
(RMSEA) of 0.08.
Subgroup comparisons were performed using χ2 tests, t-tests and one-way analyses
of variance. Pearson correlation coefficients were computed to assess bivariate
associations between variables. Statistical significance was assumed at p o 0.05.
Hierarchical multiple regression analyses were used to test the five hypotheses.
Separate analyses were performed for each of the four dependent variables: trust, work
atmosphere, emotional exhaustion and propensity to leave. In order to reduce the
number of independent variables, each of the individual background variables and
rated leader background variables were individually tested against each of the four
dependent variables (t-tests and one-way analyses of variance, except for emotional
stability and knowledge of the leader, where bivariate correlations were computed).
No statistically significant differences were found on any of the t-tests or one-way
analyses of variance, while emotional stability and knowledge of the leader both
correlated statistically significantly with each of the four dependent variables.
Following from this, it was decided to keep these two independent variables and drop
the other individual background variables and rated leader background variables.
For each of the four regression analyses a few additional participants were omitted
as they were deemed as outliers ( W 3.0 SD). Table IV shows the final number of cases
for each of the dependent variables. The independent variables were entered in the Leadership: is
following order: emotional stability; knowledge of the leader; nationality (dummy bad stronger
variables: Swiss, Swedish, Estonia, Dutch); the five destructive leadership scales; and
the constructive leadership scale.
than good?

Ethics
All participants were treated in accordance with the ethical principles of human 699
research formulated by the Swedish Research Council (2002).

Results
Dimensionality analysis of the Destrudo-L
The goodness-of-fit between two different models and the empirical outcome was
tested. The first included the five factors obtained in the original development of the
Destrudo-L (see the Methods section above). The outcome was an RMSEA of 0.082.
A second test was run where the following specified covariances between factors were
added: arrogant, unfair and threats, punishments, overdemands; passive, cowardly and
uncertain, unclear, messy; and ego oriented, false and all the other four factors.
The goodness-of-fit of this model and the empirical outcome was acceptable, an
RMSEA of 0.074 was obtained (90 per cent confidence interval 0.068-0.080).
Other common goodness-of-fit indices included a normed fit index of 0.857, a
goodness-of-fit index of 0.903 and an adjusted goodness-of-fit index of 0.876. The factor
loadings of all items but three were 0.60 or higher. In summary, the confirmative factor
analysis yielded a result which was considered as good enough to proceed with the
remaining statistical analysis using the original five-factor structure. As this SEM
analysis was not the main aim of the present study, no tables will be provided in addition
to the reported findings. Detailed results can be obtained directly from the authors.

Descriptive statistics
Table III illustrates that the destructive leadership scales generally have low means,
while the constructive leadership scale have a high mean. Two of the dependent
variables – emotional exhaustion and propensity to leave the profession – have
particularly low mean scores.
The bivariate correlations within each of the two sets of leadership scales are high
and there are strong negative correlations between the two kinds of scales.
The bivariate correlations within the four dependent variables are modest to strong.
Between these variables and the two sets of leadership scales, the bivariate correlations
are modest to strong for trust in the supervisor and work atmosphere and low
to modest for emotional exhaustion and propensity to leave. All are in the
expected direction.

Multiple regression analyses

Hierarchical multiple regression analyses were performed where the predictor
variables were regressed on each of the four dependent variables. The results are
summarized in Table IV.
Trust in the immediate supervisor. The regression equation of the final model was
statistically significant (F ¼ 105.21, p o 0.001) and the adjusted R2 was 0.69.
The destructive leadership factors explained an additional 38.7 per cent of the
variance in trust in the immediate supervisor, after controlling for emotional stability,
 37,6

700
LODJ

Table III.

deviations and
Means, standard

observed variables
correlations for the
Scale 1 2 3 4 5 6 7 8 9 10 11 12 M SD
a
1. Arrogant, unfair 0.75 0.59 0.60 0.42 0.52 −0.46 −0.45 −0.36 0.28 0.26 −0.19 −0.19 2.31 1.02
2. Threats, punishments, over-demandsa 0.69 0.58 0.44 0.49 −0.37 −0.34 −0.29 0.28 0.20 −0.19 −0.08 1.99 0.86
3. Ego oriented, falsea 0.78 0.59 0.63 −0.59 −0.52 −0.37 0.32 0.22 −0.20 −0.18 2.20 1.10
4. Passive, cowardlya 0.75 0.65 −0.58 −0.53 −0.39 0.34 0.26 −0.19 −0.16 2.20 1.06
5. Uncertain, unclear, messya 0.78 −0.63 −0.58 −0.41 0.35 0.36 −0.12 −0.18 2.15 1.04
6. Constructive leadershipa 0.96 0.72 0.52 −0.29 −0.26 0.21 0.34 4.36 0.91
7. Trust in the supervisora 0.90 0.58 −0.31 −0.27 0.16 0.33 4.27 1.23
8. Work atmospherea 0.77 −0.45 −0.38 0.12 0.27 4.69 0.81
9. Emotional exhaustiona 0.87 0.36 −0.20 −0.08 1.93 0.77
10. Propensity to leave the professiona 0.72 0.06 −0.14 1.85 0.94
11. Emotional stabilityc,d − 0.01 6.24 1.77
12. Knowledge of rated leadera,d − 3.74 1.20
Notes: n ¼ 625. aThe scale ranged from 1 to 6; bThe numbers in italic refers to Cronbach α scores; cThe scale ranged from 1 to 9; dControl variables
 Trust in the superior Work atmosphere
Leadership: is
β final model SE β final model SE bad stronger
n ¼ 525 n ¼ 525 than good?
Step 1
Emotional stability 0.001 0.018 −0.008 0.016
Adj R2 0.025 0.013
R2 change 0.022* 0.015* 701
Step 2
Knowledge of the leader 0.041 0.028 0.071 0.025
Adj R2 0.140 0.105
R2 change 0.116* 0.093*
Step 3
Nationality, Swiss 0.269* 0.100 0.195* 0.089
Nationality, Swedish 0.171* 0.081 −0.041 0.073
Nationality, Estonia 0.193 0.119 0.180 0.105
Nationality, Dutch Reference Reference Reference Reference
Adj R2 0.138 0.104
R2 change 0.003 0.005
Step 4
Arrogant, unfair −0.150* 0.045 −0.040 0.040
Threats, punishments overdemands 0.094 0.052 −0.064 0.045
Ego oriented, false −0.070 0.045 0.006 0.040
Passive, cowardly −0.080 0.042 −0.090* 0.037
Uncertain, unclear, messy −0.126* 0.047 −0.018 0.042
Adj R2 0.524 0.290
R2 change 0.387* 0.191*
Step 5
Constructive leadership 0.875* 0.054 0.323 0.050
Adj R2 0.686 0.342
R2 change 0.160* 0.053*
Total R2 0.693 0.356
Emotional exhaustion Propensity to leave
β final model SE β final model SE
n ¼ 520 n ¼ 519
Step 1
Emotional stability −0.056* 0.016 0.003 0.022
Adj R2 0.041 0.002
R2 change 0.043* 0.004
Step 2
Knowledge of the leader −0.032 0.024 −0.018 0.034
Adj R2 0.054 0.018
R2 change 0.014* 0.017*
Step 3
Nationality, Swiss −0.099 0.086 −0.119 0.124
Nationality, Swedish 0.283* 0.070 −0.141 0.100
Nationality, Estonia 0.047 0.101 0.336* 0.145
Nationality, Dutch Reference Reference Reference Reference Table IV.
Multiple regression
(continued ) analysis
LODJ Adj R2 0.069 0.037
37,6 R2 change 0.020* 0.025*
Step 4
Arrogant, unfair 0.021 0.039 0.122* 0.055
Threats, punishments overdemands 0.090* 0.044 −0.022 0.063
Ego oriented, false 0.027 0.039 −0.082 0.056
702 Passive, cowardly 0.092* 0.036 0.045 0.051
Uncertain, unclear, messy 0.103* 0.041 0.257* 0.057
Adj R2 0.191 0.155
R2 change 0.129* 0.125*
Step 5
Constructive leadership 0.040 0.045 −0.053 0.065
Adj R2 0.191 0.154
R2 change 0.001 0.001
Total R2 0.208 0.172
Table IV. Note: *Statistically significant at p o0.05

nationality and knowledge of the leader. The constructive leadership factor explained
an additional 16.0 per cent of trust in the immediate supervisor after controlling for
emotional stability, nationality, knowledge of the leader and destructive leadership
(R2 changes if entering constructive leadership before the destructive leadership behaviours:
51.7 and 3.0 per cent, respectively). In the final model, the following variables had
statistically significant β values: constructive leadership ( β ¼ 0.875, po0.001), arrogant,
unfair ( β ¼ −0.150, po0.001), nationality, Swiss ( β ¼ 0.269, po0.007), uncertain, unclear,
messy ( β ¼ 0.126, po0.008) and nationality, Swedish ( β ¼ 0.171, po0.036). H1 could
therefore be said to be supported.
Work atmosphere. The regression equation of the final model was statistically
significant (F ¼ 25.8, p o 0.001) and the adjusted R2 was 0.34. The destructive
leadership factors explained an additional 19.1 per cent of the variance in work
atmosphere, after controlling for emotional stability, nationality and knowledge of the
leader. Constructive leadership explained an additional 5.3 per cent of work atmosphere
after controlling for emotional stability, knowledge of the leader, nationality and
destructive leadership (R2 changes if entering constructive leadership behaviours
before destructive: 22.1 and 2.3 per cent, respectively). In the final model, the following
variables had statistically significant β values: constructive leadership ( β ¼ 0.323,
p o 0.001), knowledge of the leader ( β ¼ 0.071, p o 0.004), passive, cowardly
( β ¼ −0.090, p o 0.016) and nationality: Swiss ( β ¼ 0.195, p o 0.029). H2 was
therefore not supported. Since the only destructive leadership scale that gave a
significant contribution was passive, cowardly, H3 was supported.
Emotional exhaustion. The regression equation of the final model was statistically
significant (F ¼ 12.13, p o 0.001) and the adjusted R2 was 0.19. The destructive
leadership factors explained an additional 12.9 per cent of the variance in emotional
exhaustion, after controlling for emotional stability, knowledge of the leader and
nationality. Constructive leadership explained 0 per cent of emotional exhaustion after
controlling for emotional stability, knowledge of the leader, nationality and destructive
leadership (R2 changes if entering constructive leadership behaviours before
destructive: 4.9 and 8.2 per cent, respectively). In the final model, the following
variables had statistically significant β values: nationality, Swedish ( β ¼ 0.283, Leadership: is
p o 0.001), emotional stability ( β ¼ −0.056, p o 0.001), passive, cowardly ( β ¼ 0.092, bad stronger
p o 0.011), uncertain, unclear, messy ( β ¼ 0.103, p o 0.012) and threats, punishments, than good?
overdemands ( β ¼ 0.090, p o 0.041), supporting H4, that destructive leadership
behaviours show stronger positive associations with emotional exhaustion than
constructive leadership behaviours show negative associations, although the adjusted
R2 in the final model is only modestly high. 703
Propensity to leave. The regression equation of the final model was statistically
significant (F ¼ 9.59, p o 0.001) and the adjusted R2 was 0.15. The destructive
leadership factors explained an additional 12.5 per cent of the variance in work
motivation, after controlling for emotional stability, knowledge of the leader and
nationality. Constructive leadership explained an additional 0.00 per cent of work
motivation after controlling for emotional stability, knowledge of the leader, nationality
and destructive leadership (R2 changes if entering constructive leadership behaviours
before destructive: 6.5 and 6.0 per cent, respectively). In the final model, following
variables had statistically significant β values: Uncertain, unclear, messy ( β ¼ 0.257,
p o 0.001), nationality, Estonia ( β ¼ 0.336, p o 0.021) and arrogant, unfair ( β ¼ 0.122,
p o 0.029), contradicting H5.

Discussion
The aim of this study was to investigate if the hypothesis “bad is stronger than good”
also holds true for a number of leadership outcomes, more specifically trust in the
immediate leader, work atmosphere, emotional exhaustion and propensity to leave.
Based on previous research, five hypotheses were formulated. In line with the results
from Schyns and Schilling (2013), our findings are somewhat mixed. Below we discuss
our findings for each hypothesis. First, however, we conclude that the dimensional
analysis of the Destrudo-L using SEM confirmed the original five-factor structure and
justified the following analyses.
Regarding trust in the immediate supervisor, H1 was partly supported.
The correlation with trust in the immediate supervisor was higher for constructive
leadership compared to destructive leadership and it was also the constructive
leadership factor that contributed most to the final model. Concerning work
atmosphere, our results did not support H2. Passive, cowardly was the only destructive
leadership factor that gave a significant contribution to the model, supporting H3, that
the passive destructive leadership behaviours show stronger negative association with
work atmosphere than active destructive leadership behaviours. Moving on to
emotional exhaustion, besides nationality (Swedish) and emotional stability, the
destructive leadership factors passive, cowardly, uncertain, unclear, messy and threats,
punishments, overdemands gave significant contributions to the model leading to the
conclusion that H4 was supported. Finally, regarding propensity to leave
the profession, H5 was also contradicted, since the only significant contribution was
noted for nationality (Estonia) and the destructive leadership factors uncertain, unclear,
messy and arrogant, unfair. To summarize, the knowledge gained by previous research
contributed poorly to our hypotheses. In the following, we will discuss potential
explanations for our findings.
Obviously, one possible explanation for these findings may be statistical.
The factors emotional exhaustion and propensity to leave the profession had skewed
response distributions. A majority of the respondents estimated their emotional
LODJ exhaustion and their propensity to leave as low. In contrast, the factors trust in the
37,6 immediate supervisor and work atmosphere were also skewed, although not as much:
most individuals gave high ratings for trust in their leader and work atmosphere.
However, if the results are not just a statistical effect, we would also like to suggest
alternative explanations for our findings. First, a closer look at our four dependent
variables reveals a certain pattern. Constructive leadership behaviours covaried more
704 strongly with trust in the immediate supervisor and work atmosphere, while destructive
leadership behaviours were more strongly associated with emotional exhaustion and
propensity to leave. Trust in the immediate leader and work atmosphere can be regarded
as positive phenomena while emotional exhaustion and propensity to leave can be
considered as negative phenomena. The first two are positively formulated and the two
latter negatively. These two pairs of phenomena also have another common
denominator: trust in the immediate supervisor and work atmosphere have an
external focus and are work-related concepts, i.e. the items mainly focus on others who
are significant in one’s working life. Likewise, emotional exhaustion and propensity to
leave express a more personal meaning, i.e. they estimate the individual’s feelings and
intentions. This suggests that constructive leadership behaviours possibly have a greater
impact on positive phenomenon and/or phenomenon associated with work-related
relationships. On the other hand, destructive leadership behaviours appear to have a
greater impact on negative phenomena with a stronger personal meaning, which is in line
with previous studies. For example, Schyns and Schilling (2013) found that destructive
leadership showed higher negative correlations with commitment and well-being while
constructive leadership had higher positive correlations with attitude towards the leader
and job satisfaction.
The responses on the positive outcome variables were explained to a higher degree
by the predictor variables compared to the negatively formulated outcome variables.
Thus, 69 per cent of trust in the immediate supervisor and 34 per cent of experienced
work atmosphere was explained, compared to only 19 per cent on emotional exhaustion
and 15 per cent on propensity to leave the profession. This shows that more
person-oriented factors are mainly explained by other aspects. It also indicates the
leader’s limitations regarding influence on these aspects. They are mainly explained by
other phenomena that are not controlled for in this study. For instance, trust in leaders
have been suggested to be more determined by the leader’s behaviours than
demographics, personality and structural factors (Dirks, 2006), while research supports
that emotional exhaustion is not only influenced by leadership but also by other
aspects such as personal coping strategies, emotional culture and personal resources
(Grandey et al., 2005; Ito and Brotheridge, 2003).
Our results show that the passive forms of destructive leadership had a stronger
impact on our investigated dependent variables compared to the active forms, except
for trust in the immediate supervisor. One reason for why arrogant, unfair gave the
strongest contribution, of the destructive leadership factors, to trust in in the immediate
supervisor may be related to the fact that the factor can be said to express a lack of
both integrity and benevolence which has been suggested to be core components of
trust in leader (together with ability) (Mayer et al., 1995). Fairness (or lack of) has also
been shown to be associated with trust in leaders (Dirks and Ferrin, 2002).
Elaborating on why the passive forms appears to have a stronger impact on the
dependent variables, it can be argued that these behaviours most likely are intermittent
and therefore easier for the individual to deal with and have less effect on the daily
work. A leader who behaves arrogantly, is unfair and/or uses threats may have the
greatest negative effect on subordinates when he or she is present. In contrast, passive Leadership: is
behaviours can be considered as reflecting a more lasting attitude or having a more bad stronger
lasting effect on subordinates’ work climate. Passive behaviours have been found to be
more likely to cause frustration and problems (even when the leader is not present) in
than good?
terms of the leader creating a work climate characterized by uncertainty, role
ambiguity and conflicts (Einarsen, 1999; Leymann, 1996; Skogstad et al., 2007). These
behaviours are probably more subtle and difficult to detect. One could argue that 705
subordinates who score low on trust and high on emotional exhaustion and propensity
to leave are more inclined to rate the leader’s behaviour as destructive. However,
emotional stability is controlled for in step 1 of the regression analysis, indicating that
it is not emotionally instable individuals who account for these results.
Speculating on possible consequences of these findings, we would like to point to a
potential problem. The factors passive, cowardly and uncertain, unclear, messy, capture
behaviours related to the leader being absent, not showing active interest, giving unclear
instructions and behaving confused. These are behaviours that may be consequences of
leader-related antecedents such as work-related stress. Due to shortage of time, leaders
therefore may behave destructively and since one of our largest work life problems is
stress, this implies that many leaders who would normally not be prone to behaving
destructively, may fall into using passive destructive leader behaviours with
unfavourable consequences for subordinates’ health and performance.
Like previous results, our findings show modest to strong correlations between the
different destructive leadership factors and constructive leadership supporting the
finding that destructive and constructive leadership behaviours do not exist apart but
reflect integrated parts of leadership (Aasland et al., 2010). In order to understand leaders’
influence on different kinds of work- and organizational-related consequences, both
destructive and constructive leadership behaviours need to be taken into consideration.
Finally, regarding practical implications, our results emphasize the importance of
focusing on both constructive and destructive leadership at the selection stage, as well
as during training of military leaders. Focusing on them separately obstructs optimal
leader development and prevents leaders from gaining authentic self-knowledge. The
results also point at the importance of including both aspects of leadership in leader
evaluation processes.
Our results indicate further that destructive leadership has a stronger impact on
negative phenomena with a stronger personal meaning. Krasikova et al. (2013) suggest
that organizations are more prone to intervene when destructive leaders encourage
subordinates to pursue destructive goals than when they use destructive methods of
influence with followers. This can be valuable information for leaders as well as experts
who work in human resources. Destructive leaders can have severe negative impact on
individuals, leading to increased sick leave and/or turnover which, in turn, results in
negative consequences for the organization. Passive destructive behaviours may
perhaps be more difficult to discover compared to active forms which may lead to a less
chance of someone detecting and intervening. In this respect, it is important to create
structures for how the organization can detect and handle destructive leaders.

Limitations and future outlook

One study limitation is related to item construction. Two of the dependent variables were
formulated positively (trust in the immediate supervisor, work atmosphere) and two were
negatively formulated (emotional exhaustion, propensity to leave) and perhaps
the discovered patterns would have changed if the items were formulated uniformly.
LODJ A second weakness is the possibility of common method variance and response set
37,6 tendencies (Podsakoff et al., 2003) since all data were collected using self-report
questionnaires. The scales need to be tested with other actors as raters, and ideally, also
using more objective outcome measures. Following from this, the results need to be
interpreted with caution.
A third weakness is connected to the translation of the instruments used in present
706 study. When translating an instrument to new languages a translation/back translation
procedure should be followed in order to secure that the instrument is conceptually
equivalent in each of the participating countries. This procedure was followed when
the Destrudo-L and DLQ was translated from Swedish to English. However, the
authors have no knowledge about the translation process in the other participating
countries and therefore there is a risk that the concepts are being interpreted differently
by the participants.
A final limitation is related to the lack of response rate for the Swiss data, reducing
the generalizability of the data. Caution should be exercised in this case.
Finally, we would like to make some suggestions for further studies. To find out if
bad or good is stronger regarding leadership, more studies need to be conducted in
order to investigate the circumstances under which bad prevails over good and
vice versa. One suggestion is to use measurements that are formulated in the same way,
i.e. with the dependent variable measures all either positively or negatively formulated.
Regarding antecedents, narrower bandwidth traits should be tested in addition to the
general Big Five dimension used in this study. The predictive validity of carefully
selected narrower traits has been shown to be higher compared to broader constructs
(Ashton et al., 1995; Bilgiç and Sümer, 2009). It has, for instance, been reported that two
facets of the neuroticism dimension – anxiety and anger – covary in opposite directions
with sub forms of destructive leadership behaviours (Kant et al., 2013).
The fact that the passive forms of destructive leadership had the strongest impact
also deserves further studies. Our study is cross-sectional and it may be possible that
passive forms of destructive leadership are damaging in the long run while active
forms are immediately damaging – to use the words of Thoroughgood et al. (2012).
We therefore suggest that future studies examine the longitudinal association between
different forms of destructive leadership and outcomes such as trust, emotional
exhaustion and propensity to leave.
Leadership is assumed to be context-sensitive (Osborn et al., 2002). Therefore we
suggest that future research attempts to replicate this study in other contexts, for
example, in civilian contexts or during international military operations. Previous studies
show that leaders who are good at structuring and who take prompt and decisive action
are seen as more effective during critical incidents (Flanagan and Levy, 1952, cited in
Hannah et al., 2010). It can be argued that passive destructive leadership behaviours
should have the strongest negative consequences during international operations where
passive behaviours leading to confusion and role ambiguity, can have deadly outcomes.

References
Aasland, M., Skogstad, A., Notelaers, G., Birkeland Nielsen, M. and Einarsen, S. (2010),
“The prevalence of destructive leadership behaviour”, British Journal of Management,
Vol. 21 No. 2, pp. 438-452.
Ashforth, B.E. (1994), “Petty tyranny in organizations”, Human Relations, Vol. 47 No. 7,
pp. 755-778.
Ashton, M.C., Jackson, D.N., Paunonen, S.V., Helmes, E. and Rothstein, M.G. (1995), “The criterion Leadership: is
validity of broad factor scales versus specific facet scales”, Journal of Research in
Personality, Vol. 29 No. 4, pp. 432-442.
bad stronger
than good?
Baumeister, R.F., Bratslavsky, E., Finkenauer, C. and Vohs, K.E. (2001), “Bad is stronger than
good”, Review of General Psychology, Vol. 5 No. 4, pp. 323-370.
Belenky, G.L., Noy, S.L. and Solomon, Z. (1985), “Battle stress: the Israeli experience”,
Military Review, Vol. 65 No. 7, pp. 28-37. 707
Bilgiç, R. and Sümer, H.C. (2009), “Predicting military performance from specific personality
measures: a validity study”, International Journal of Selection and Assessment, Vol. 17
No. 2, pp. 231-238.
Burris, E.R., Detert, J.R. and Chiaburu, D.S. (2008), “Quitting before leaving: the mediating effects
of psychological attachment and detachment on voice”, Journal of Applied Psychology,
Vol. 93 No. 4, pp. 912-922.
Bycio, P., Hackett, R.D. and Allen, J.S. (1995), “Further assessments of Bass’s (1985)
conceptualization of transactional and transformational leadership”, Journal of Applied
Psychology, Vol. 80 No. 4, pp. 468-478.
Campbell, D.J. (2012), “Leadership in dangerous contexts”, in Laurence, J.H. and Matthews, M.D.
(Eds), The Oxford Handbook of Military Psychology, Oxford University Press, New York, NY,
pp. 158-175.
Chen, H.C. and Kao, H.S.R. (2009), “Chinese paternalistic leadership and non-Chinese
subordinates’ psychological health”, International Journal of Human Resource
Management, Vol. 20 No. 12, pp. 2533-2546.
Chi, S.-C.S. and Liang, S.-G. (2013), “When do subordinates’ emotion-regulation strategies matter?
Abusive supervision, subordinates’ emotional exhaustion, and work withdrawal”,
The Leadership Quarterly, Vol. 24 No. 4, pp. 125-137.
Clements, C. and Washbush, J.B. (1999), “The two faces of leadership: considering the dark side of
leader-follower dynamics”, Journal of Workplace Learning, Vol. 11 No. 5, pp. 170-175.
Collins, J.J. and Jacobs, T.O. (2002), “Trust in the profession of arms”, in Snider, D.M. and
Watkins, G.L. (Eds), The Future of the Army Profession, The McGraw-Hill, Boston, MA,
pp. 39-58.
Costa, P.T. Jr and McCrae, R.R. (1992), NEO PI-R: Professional Manual, Psychological
Assessment Resources, Odessa, FL.
Craig, B. and Kaiser, R.B. (2012), “Destructive leadership”, in Rumsey, M.D. (Ed.), The Oxford
Handbook of Leadership, Oxford University Press, Oxford, NY, pp. 439-454.
Dirks, K.T. (2006), “Three fundamental questions regarding trust in leaders”, in Bachman, R.
and Zaheer, A. (Eds.), Handbook of Trust Research, MPG Books Group, pp. 15-28.
Dirks, K.T. and Ferrin, D.L. (2002), “Trust in leadership: meta-analytic findings and implications
for research and practice”, Journal of Applied Psychology, Vol. 87 No. 4, pp. 611-628.
Duffy, M.K., Ganster, D.C. and Pagon, M. (2002), “Social undermining in the workplace”, Academy
of Management Journal, Vol. 45 No. 2, pp. 331-351.
Einarsen, S. (1999), “The nature and causes of bullying at work”, International Journal of
Manpower, Vol. 20 Nos 1/2, pp. 16-27.
Einarsen, S., Aasland, M. and Skogstad, A. (2007), “Destructive leadership behaviour: a definition
and conceptual model”, The Leadership Quarterly, Vol. 18 No. 3, pp. 207-216.
Fors Brandebo, M. and Larsson, G. (2012), “Influence of IED attacks on leadership: dealing with
the invisible enemy”, Res Militaris, Vol. 2 No. 3.
LODJ Frischer, J. and Larsson, K. (2000), “Laissez-faire in research education – an inquiry into a
Swedish doctoral program”, Higher Education Policy, Vol. 13 No. 2, pp. 132-155.
37,6
Glasø, L., Einarsen, S., Matthiesen, S.B. and Skogstad, A. (2010), “The dark side of leaders:
a representative study of interpersonal problems among leaders”, Scandinavian Journal of
Organizational Psychology, Vol. 2 No. 2, pp. 3-14.
Grandey, A.A., Fisk, G.M. and Steiner, D.D. (2005), “Must ‘service with a smile’ be stressful?
708 The moderating role of personal control for American and French employees”, The Journal
of Applied Psychology, Vol. 90 No. 5, pp. 893-904.
Hannah, S.T., Campbell, D.J. and Matthews, M.D. (2010), “Advancing a research agenda for
leadership in dangerous contexts”, Military Psychology, Vol. 22 No. S1, pp. 157-189.
Harvey, P., Stoner, J., Hochwarter, W. and Kacmar, C. (2007), “Coping with abusive supervision:
the neutralizing effects of ingratiation and positive affect on negative employee outcomes”,
The leadership Quarterly, Vol. 18 No. 3, pp. 264-280.
Hinkin, T.R. and Schriesheim, C.A. (2008), “An examination of ‘nonleadership’: from laissez-faire
leadership to leader reward omission and punishment omission”, Journal of Applied
Psychology, Vol. 93 No. 6, pp. 1234-1248.
Ito, J.K. and Brotheridge, C.M. (2003), “Resources, coping strategies, and emotional exhaustion:
a conservation of resources perspective”, Journal of Vocational Behavior, Vol. 63 No. 3,
pp. 490-509.
Jones, E., Massey Kantak, D., Futrell, C.M. and Johnston, M.W. (1996), “Leader behavior,
work-attitudes, and turnover of salespeople: an integrative study”, Journal of Personal
Selling & Sales Management, Vol. 16 No. 2, pp. 13-23.
Kant, L., Skogstad, A., Torsheim, T. and Einarsen, S. (2013), “Beware of the angry leader: trait
anger and trait anxiety as predictors of petty tyranny”, The Leadership Quarterly, Vol. 24
No. 1, pp. 106-124.
Kelloway, E.K., Sivanathan, N., Francis, L. and Barling, J. (2005), “Poor leadership”, in Barling, J.,
Kelloway, E.K. and Frone, M.R. (Eds), Handbook of Work Stress, Sage Publications,
Thousand Oaks, CA, pp. 89-112.
Kolditz, T.A. (2007), In Extremis Leadership: Leading as if Your Life Depended on it, Jossey-Bass,
San Francisco, CA.
Krasikova, D.V., Green, S.G. and LeBreton, J.M. (2013), “Destructive leadership: a theoretical
review, integration, and future research agenda”, Journal of Management, Vol. 39 No. 5,
pp. 1308-1338.
Larsson, G. (2006), “The developmental leadership questionnaire (DLQ): some psychometric
properties”, Scandinavian Journal of Psychology, Vol. 47 No. 4, pp. 253-262.
Larsson, G., Fors Brandebo, M. and Nilsson, S. (2012), “Destrudo-L: development of a short scale
designed to measure destructive leadership behaviours in a military context”, Leadership
& Organization Development Journal, Vol. 33 No. 4, pp. 383-400.
Larsson, G., Fors, M., Levin, A.-L. and Thuresson, A. (2007), Ledarskapets betydelse för värnpliktigas
och unga officerares motivation att fortsätta inom Försvarsmakten (The Importance of
Leadership to the Motivation of Conscripts and Young Officers to Continue in the Armed
Forces) (ILM Serie I:38), Försvarshögskolan, Institutionen för ledarskap och management,
Karlstad.
Larsson, G., Carlstedt, L., Andersson, J., Andersson, L., Danielsson, E., Johansson, A., Johansson, E.,
Robertsson, I. and Michel, P.-O. (2003), “A comprehensive system for leader evaluation and
development”, Leadership & Organization Development Journal, Vol. 24 No. 1, pp. 16-25.
Leymann, H. (1996), “The content and development of mobbing at work”, European Journal of
Work and Organizational Psychology, Vol. 5 No. 2, pp. 165-184.
Maslach, C., Jackson, S.E. and Leiter, M.P. (1996), Maslach Burnout Inventory Manual, 3rd ed., Leadership: is
Consulting Psychologists Press, Palo Alto, CA.
bad stronger
Mayer, R.C., Davis, J.H. and Schoorman, F.D. (1995), “An integrative model of organizational than good?
trust”, Academy of Management Review, Vol. 20 No. 3, pp. 709-734.
Mullen, J., Kelloway, E.K. and Teed, M. (2011), “Inconsistent style of leadership as a predictor of
safety behaviour”, Work & Stress, Vol. 25 No. 1, pp. 41-54.
Ness, J., Jablonski-Kaye, D., Obigt, I. and Lam, D.M. (2011), “Understanding and managing stress”, 709
in Sweeney, P.J., Matthews, M.D. and Lester, P.B. (Eds), Leadership in Dangerous
Situations. A Handbook for the Armed Forces, Emergency Services, and First Responders,
Naval Institute Press, Annapolis, MD, pp. 40-59.
Neuman, J.H. and Baron, R.M. (2005), “Aggression in the workplace: a social-psychological
perspective”, in Fox, S. and Spector, P.E. (Eds), Counterproductive Work
Behavior: Investigations of Actors and Targets, American Psychological Association,
Washington, DC, pp. 13-40.
Nilsson, S., Jonsson, E., Fors Brandebo, M. and Larsson, G. (2012), “Military professionalism of the
Swedish armed forces in the 21st century”, in Stouffer, J. and Lindsay, D. (Eds), Threats to
Military Professionalism: International Perspectives, Canadian Defence Academy Press,
Kingston, pp. 169-194.
Osborn, R.N., Hunt, J.G. and Jauch, L.R. (2002), “Toward a contextual theory of leadership”,
The Leadership Quarterly, Vol. 13 No. 6, pp. 797-837.
Podsakoff, P.M., MacKenzie, S.B., Lee, J.Y. and Podsakoff, N.P. (2003), “Common method biases in
behavioral research: a critical review of the literature and recommended remedies”, Journal
of Applied Psychology, Vol. 88 No. 5, pp. 879-903.
Reed, G.E. and Bullis, R.C. (2009), “The impact of destructive leadership on senior military officers
and civilian employees”, Armed Forces and Society, Vol. 36 No. 1, pp. 5-18.
Schyns, B. and Schilling, J. (2013), “How bad are the effects of bad leaders? A meta-analysis
of destructive leadership and its outcomes”, The Leadership Quarterly, Vol. 24 No. 1,
pp. 138-158.
Skogstad, A., Hetland, J., Glasø, L. and Einarsen, S. (2014), “Is avoidant leadership a root cause of
subordinate stress? Longitudinal relationships between laissez-faire leadership and role
ambiguity”, Work & Stress, Vol. 28 No. 4, pp. 323-341.
Skogstad, A., Einarsen, S., Torsheim, T., Aasland, M.S. and Hetland, H. (2007), “The destructiveness
of laissez-faire leadership behavior”, Journal of Occupational Health Psychology, Vol. 12 No. 1,
pp. 80-92.
Swedish Research Council (2002), “Research ethical principles in humanistic-social research”,
Vetenskapsrådet, Stockholm.
Sweeney, P.J. (2010), “Do soldiers re-evaluate trust in their leaders prior to combat operations?”,
Military Psychology, Vol. 22 No. 1, pp. 70-88.
Sweeney, P.J., Matthews, M.D. and Lester, P.B. (2011), “Leading in dangerous situations:
an overview of the unique challenges”, in Sweeney, P.J., Matthews, M.D. and Lester, P.B.
(Eds), Leadership in Dangerous Situations: A Handbook for the Armed Forces,
Emergency Services, and First Responders, Naval Institute Press, Annapolis, MD,
pp. 3-18.
Sweeney, P.J., Thompson, V. and Blanton, H. (2009), “Trust and influence in combat:
a interdependence model”, Journal of Applied Social Psychology, Vol. 39 No. 1, pp. 235-264.
Tepper, B.J. (2000), “Consequences of abusive supervision”, Academy of Management Journal,
Vol. 43 No. 2, pp. 178-190.
LODJ Thoroughgood, C.N., Tate, B.W., Sawyer, K.B. and Jacobs, R. (2012), “Bad to the bone: empirically
defining and measuring destructive leader behavior”, Journal of Leadership &
37,6 Organizational Studies, Vol. 19 No. 2, pp. 230-255.
Woods, S.A. and Hampson, S.E. (2005), “Measuring the Big Five with single items using a bipolar
response scale”, European Journal of Personality, Vol. 19 No. 5, pp. 373-390.
Yukl, G. (2006), Leadership in Organizations, Pearson Prentice Hall, Upper Saddle River, NJ.
710
Further reading
Tepper, B.J. (2007), “Abusive supervision in work organizations: review, synthesis and research
agenda”, Journal of Management, Vol. 33 No. 3, pp. 261-289.

About the authors

Maria Fors Brandebo (PhD, Karlstad University, Sweden) is a Researcher at the Department of
Security, Strategy and Leadership, Swedish Defence University, Sweden. She has published
articles, book chapters and research reports within the field of psychology and leadership,
motivation, trust and organization. Maria Fors Brandebo is the corresponding author and can be
contacted at: [email protected]
Sofia Nilsson (PhD, Karlstad University, Sweden) is a Researcher at the Department of
Security, Strategy and Leadership, Swedish Defence University, Sweden. She has published
several journal articles, chapters in books and research reports at the Swedish Defence University
dealing with management, leadership, organization and stress in general, and moral stress in
particular.
Gerry Larsson (PhD, Gothenburg University, Sweden) is a Licensed Psychologist and
Professor of Leadership Psychology at the Swedish Defence University. He is also Professor of
Stress Psychology at the Hedmark University College, Norway. During the period 2004-2009 he
served as Vice President of the Swedish Defence University. He has published over 130 journal
articles, 80 books or chapters in books and more than 560 research reports.

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: [email protected]